Transform 3-Point estimates to normal form
Registered by
Michael-Olaf
The idea is that the input is a 3-Point estimate, but the random number distribution is in a normal form. A transformation is needed from this 3-Point estimate to the 2 variables needed for normal distribution
Blueprint information
- Status:
- Not started
- Approver:
- Michael-Olaf
- Priority:
- Medium
- Drafter:
- Michael-Olaf
- Direction:
- Approved
- Assignee:
- Michael-Olaf
- Definition:
- Approved
- Series goal:
- Accepted for trunk
- Implementation:
-
Not started
- Milestone target:
-
2.0
- Started by
- Completed by
Whiteboard
Proposal 1:
Input variables are:
o1 = optimistic
m1 = most likely
p1 = pessimistic
1st step: compute median and a deviation:
median = (o1+4*m1+p1)/6
deviation = (o1+p1)/2 - o1
2nd step compute new input variables for random number with normal distribution:
o2 = max(median-
p2 = median+deviation
Examples:
o1=50, m1=60, p1=70 => o2=50, p2=70
o1=50, m1=55, p1=70 => o2=46.667, p2=66.667
o1=5, m1=6, p1=25 => o2=1, p2=19
(?)
Work Items
Dependency tree
* Blueprints in grey have been implemented.
